Project Description
Third-cycle subject: Computer Science or Speech and Music Communication
We are looking for a doctoral student interested in Artificial Intelligence, Machine Learning, Natural Language Processing and Human-Machine Interaction, that will work in a newly funded project at the Department of Speech, Music and Hearing (www.kth.se/tmh) within the School of Electrical Engineering and Computer Science at KTH. The project is financed by WASP which offers a graduate school with research visits, partner universities, and visiting lecturers.
The newly started project is titled “Representation learning for Conversational AI”, and is a collaboration project between KTH Royal Institute of Technology in Stockholm and Chalmers in Gothenburg. The aim of the project is to develop generic language models for spoken conversation. Whereas traditional language models in NLP (such as BERT) have focused on written language, we want to model spoken conversation, where aspects such as speech acts, turn-taking, incremental processing and prosody are of importance. This means that the models will have to process both text and audio (and potentially other modalities). In collaboration with industry, we will then explore how such models can be used for applications such as conversational systems, voice assistants and social robots. Another important focus will be on model analysis and visualization.
More information about the project can be found here: https://www.kth.se/profile/skantze/page/phd-student-position-representation-learning-for-conversational-ai
Supervision: The doctoral student will be supervised by Prof. Gabriel Skantze